Covering the spread

Simon Jackman at The Bullring has done as I did in an idle moment a few weeks ago, deriving various measures of spread from an election’s worth of swing results at national and state level. If Jackman’s interpretation is correct, a state-level poll is little better than a national one as a pointer to a given seat result:

We see that there is less variation in swing within states than there is overall. But not that much less. The measures of spead of the swings (range, standard deviations, the mean absolute deviation around the mean) for each state are still quite large relative to corresponding national figure … There is still an awful lot of variability in swings out there, and I’d be reluctant to start applying uniform swing models within states.

However, he does add “just one caveat to all of this”:

It could well be that when average swings are large (or dare I say massive), there is greater uniformity or even less uniformity than when average swings are relatively small. I haven’t looked at data from previous election to know the answer to that, but it be helpful to know the answer to that.

Which is easily done if you have a spreadsheet full of swing figures, like I do. These three tables replicate Jackman’s for the 1996, 1998 and 2001 elections, two of which saw heavy traffic from one party to the other.

2001 Nat’l NSW Vic Qld SA WA
Mean 1.9 3.1 1.4 2.1 0.2 1.2
SD 2.4 2.8 2.0 2.4 2.3 1.4
MAD 1.8 2.2 1.5 1.7 2.1 1.0
Minimum -5.5 -4.0 -5.5 -3.3 -2.7 -2.4
Maximum 10.1 10.1 4.7 7.9 3.3 3.3
Range 15.6 14.1 10.2 11.2 6.0 5.7
N 142 45 37 25 11 15
1998 Nat’l NSW Vic Qld SA WA
Mean -4.8 -4.5 -3.2 -7.1 -4.0 -6.2
SD 2.9 2.2 2.3 3.2 2.8 2.7
MAD 2.3 1.9 1.8 2.4 2.2 1.7
Minimum -15.3 -10.2 -9.8 -15.3 -9.0 -11.1
Maximum 0.3 -0.3 0.3 -0.5 -0.2 0.0
Range 15.6 9.9 10.1 14.8 8.8 11.1
N 139 48 35 25 11 12
1996 Nat’l NSW Vic Qld SA WA
Mean 5.2 7.0 1.7 8.3 4.5 2.2
SD 3.3 2.4 1.4 2.5 1.3 1.4
MAD 2.7 2.0 1.1 2.0 1.0 1.2
Minimum 0.1 2.3 0.1 4.4 2.2 0.2
Maximum 14.0 12.0 4.9 14.0 6.5 3.9
Range 13.9 9.7 4.8 9.6 4.3 3.7
N 138 48 34 25 12 11

And what do you know. The last time there was a big swing and a change of government, the gap between measures of spread at state and national level was significantly higher than in 2004. However, this was not true of the 1998 election, which saw a substantial swing to Labor but no change of government. That might be due to the effect of One Nation in polarising the cities and the regions, most evidently in Queensland. The even messier picture from 2001 provides support for Jackman’s suggestion that a relative lack of state-level uniformity might be a phenomenon of status quo elections.

UPDATE: Geoff Lambert, who knows way more about these things than I do, offers a well-made point about the leptokurticity (here, use my hankie) of swing distributions in comments.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

587 comments on “Covering the spread”

Comments Page 1 of 12
1 2 … 12

  2. Umm.. that was fun. How many elections (sets of data) have we got to play with?

    Narrator: There is a theory which states that if ever anyone discovers exactly what the Universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarrely inexplicable.

  4. Nicely done William. I think I have the data lying around here somewhere, but enuf was enuf for that Bullring post. There is an important point here: if the swing is on, then maybe its on, and those state level aggregations have a little more predictive efficacy.

    In 1996 NSW and Qld are the wildcards, so to speak, with remarkable within-state uniformity in Vic, SA and WA. Curious…

  5. An analysis of electorate swings (1972-2004, N= about 1,800) produces a near normal distribution, but slightly “leptokurtic”, indicating that swings tend to cluster rather too closely about the mean, which is 0.5%. The standard deviation of the swing is 4.7% overall, but, when analysed on a single-election basis can be shown to be dependant upon the Australia-wide swing, rising with the latter from about 2.6% where there is no nett swing, to 5.3% where the swing is 7.5% (the highest national swing ever seen in the 1972-2004 period).

    The pendulum, that popular weapon of the psephologist is a very useful predictive device at election time. It is admitted by its proponents that swings vary from electorate to electorate but they argue that any “errors” arising from this tend to cancel out, meaning that the overall prediction is still likely to be “nearly right”. Actually, the pendulum has proven to be a rather poor predictor of nett seat change when the swing and potential seat change is small (less than 10 seats). In general, the performance in both a relative and an absolute sense increases rapidly above a 10 seat change. Knowing the statistics for means and standard deviations of swings, one can conduct a simulation to model the pendulum’s performance. For the current election with a swing of (say) 7%, where the expected seat change is about 30, repeated simulations show that the nett seat changes could be 5 above or below this in 95% of such elections.

  8. According to RBA Website

    7 November 2007
    Outcome of Board Meeting (9.30 am)

    12 November 2007
    Statement on Monetary Policy (11.30 am) – hopefully another hit on the govt next week

    You can even sign up for an email service for their latest news – oooh! 😉

  14. Ave it, don’t unneedlessly give your cred a hit. It’s pennies to a pound to go up, and I can bet you the entire Atlantic Ocean it won’t go down.

  16. Yeah I dont get it, anyway, Albrechtsen in the Australian today wirtes that Gillard is second rate. I wrote in her column and reproduce it here as she wont publish it
    ‘Second rate? What does that make you Janet?’

    Airheadsen is a really nasty piece of I dont know what. Gillard will be PM one day and Airheadsen will always be a third world tabloid journo.

    I thank you.

  17. 15 – I’m an optimist! Even tories in the mother country staying up late to catch news of the whumping of their brethren down under have cred in my eyes.

  19. Channel Nine news this morning (ie 8:00 am in Victoria) showing a stat from the Australian’s newspoll showing a 10 point change in the response to “who does the economy better” > Kevin Rein is now down 5% to 32% on that question.

    I presume this was part of the last round of Newspoll on Monday, but I haven’t seen it anywhere.

    Has anyone seen it?

  20. Did anyone listen to Radio National today after 6.30am? 2 articles one on Rudd and the other on Howard. A more unbalanced couple of reports would be hard to find. Essentially Rudd is only tolerated in Qld and basically ate babies when he worked for Goss vs a PM who is laughing merrily and seems far from concerned about a possible loss.

    That for me is the end of RN Breakfast and Fran Kelly. I will now rely on blogs for getting the news links.

  22. In my view, asking voters ‘who can best manage the economy’ is a useless polling question because the Opposition have never had a go! How can a voter assess the likely performance of an alternative government in ANY area of policy? When asked about the economy, voters will always lean towards the government because they have seen the evidence.

  23. 20

    I agree. Kelly & Triolli, both lightweight no-nothings. I doubt Kelly has much editorial input though. Don’t know about Triolli – only get her on Friday Lateline. Used to be my ‘must see’ edition but since Virginia it’s not worth a look.

  24. As Sol Lebovic has had a column in the Oz today doing a Bludger Spread Analysis on a state basis, here are the actual figures for the 1972-2004 elections for the mean swing and the standard deviation of it.

    STATE SWING STD DEV
    ACT -0.6% 4.8%
    NSW -0.8% 4.8%
    NTE 0.1% 2.9%
    QLD -1.4% 4.4%
    SAU -1.0% 3.7%
    TAS -0.4% 5.4%
    VIC 0.4% 4.5%
    WAU -0.7% 5.7%

    You can’t read anything into the means, but there are significant differences in the SD’s- Western Australia is more volatile than Victoria, for instance.

  27. Just something for our conservative friends to munch on from that bastion of conservatism Sol Lebovic – “If present polling levels hold through to the election though, not much analysis of state differences will matter with Labor romping in.”

  29. The ABC: useless currently! You’re correct, Fran Kelly/Viriginia Trioli/Chris Ullmann are all running a very anti-Rudd line, presumably on the instruction of the ABC board. Radio is worse than TV. Kerry O’Brien obviously hates Howard, and Jim Middleton’s TV report last night was a favourable one for Labor.

  30. [20] – Yeah RN has been quite strange during this campaign. They seem to have a journalist on the road with “Team Howard”, but none with Rudd. What is that about?

    Fran seems to dislike Rudd in quite a personal way – she can be critical of Howard’s policies, but her approach to Rudd is quite personal.

    Weird.

  31. “The ABC: useless currently! You’re correct, Fran Kelly/Viriginia Trioli/Chris Ullmann are all running a very anti-Rudd line, presumably on the instruction of the ABC board. Radio is worse than TV. Kerry O’Brien obviously hates Howard, and Jim Middleton’s TV report last night was a favourable one for Labor.”

    Heard Uhlman interview Andrew RObb this morning on interest rates and the economy, and he was pretty hard on him.

    Even got to the point where he said “What is the coalition message? You’re standing behind posters saying go for growth, yet you’re now saying everyone should be scared of inflation. What are you trying to tell the Australian people?”

    Was pretty good actually.

    PS: Have tried to do a quote with the 1st para. If it doesn’t work, can someone enlighten me on it please?

  32. “Fran seems to dislike Rudd in quite a personal way”

    Maybe it’s because she’s gay and Rudd has shown himself to be no friend of gays.

  33. Thanks to Howard’s board appointments ABC is slowly turning into a refuge for second rate right wing journos. Give PM another 3 years and all the voices of reason and balance at ABC will be replaced by opinion writers from tabloids!

  34. “An online survey conducted for The Australian has found that less than 20 per cent of respondents would blame Mr Howard if rates go up. A Galaxy poll in Monday’s Daily Telegraph found that only 12 per cent of voters would blame the PM – and that figure only rose to 20 per cent among Labor voters.”

    I participated in that poll and I give it little credibility. The way the questions were couched was absolutely jaw dropping. Blatant push polling.

    If you were at all fair minded it was going to come out in the Coalition’s favour.

  36. Not to mention the fact they are asking the wrong question. I don’t particularly blame the Coalition for recent interest rate rises, but neither do I think they deserved credit in 2004 for keeping rates low. The issue is one of dishonesty and trying to have it both ways.

    You can’t take the credit when things are going well and then distance yourself from responsibility when they go bad. And that’s the point Rudd has to hammer.

  37. Ross Gittins in the SMH bells the cat on the way interest rates and the economy works:

    Why are the people who make their living tracking and predicting movements in interest rates so unconcerned about which side wins the election? Because they know that elected governments have little influence over interest rates and the management of the macro economy.

    It’s clear from the way Howard and Costello have had to change their tune during the course of this campaign that, far from being able to influence interest rates, they didn’t even see this rise coming. As long ago as August, the Reserve Bank governor, Glenn Stevens, warned that if the consumer price index figures due on October 24 revealed a continued build-up of inflation pressure, he would have little choice but to raise rates, even if an election campaign were in progress at the time.

    So Howard could have called the election well before yesterday’s meeting of the Reserve Bank board, but he didn’t bother. That may go down as one of the great miscalculations of modern politics.

    Rather, he chose a campaign slogan, Go for Growth, that’s now proved embarrassingly inappropriate and has had to be ditched. He wanted to claim the credit for the economy’s rapid growth and promise that under the Libs it could continue indefinitely, leading us back to full employment.

    Whoops. It’s the rapid growth that’s making the Reserve so nervous about inflation and will have prompted it to raise interest rates twice in the past three months. Its objective is clear: to slow the rate of growth and thereby ease inflation pressure.

    A fortnight ago Howard and Costello were claiming the inflation rate was at its lowest in almost nine years. Now they say there’s a lot of inflation in the system so only they can be trusted to manage this “more challenging and difficult economic outlook”.

    Will the punters fall for this two elections in a row? I doubt it. But Labor has just as much interest as the Libs in maintaining the delusion that politicians run the economy. Don’t believe either side.

    http://www.smh.com.au/news/opinion/the-treasurer-is-mainly-pretence/2007/11/06/1194329223119.html

  43. 32
    Spiros Says:
    November 7th, 2007 at 9:15 am
    “Fran seems to dislike Rudd in quite a personal way”

    Maybe it’s because she’s gay and Rudd has shown himself to be no friend of gays.

    And Howard is??

  50. Just listening to 3AW. You’d think by those calling in that this rate rise will have no impact whatsoever. I don’t believe it personally.

Comments are closed.

Comments Page 1 of 12
1 2 … 12